Portal de Periódicos da CAPES

UNIQ

2019; Association for Computing Machinery; Volume: 37; Issue: 1-4 Linguagem: Inglês

10.1145/3444943

1557-7333

Chaim Baskin, Natan Liss, Eli Schwartz, Evgenii Zheltonozhskii, Raja Giryes, Alex Bronstein, Avi Mendelson,

Adversarial Robustness in Machine Learning

We present a novel method for neural network quantization. Our method, named UNIQ , emulates a non-uniform k -quantile quantizer and adapts the model to perform well with quantized weights by injecting noise to the weights at training time. As a by-product of injecting noise to weights, we find that activations can also be quantized to as low as 8-bit with only a minor accuracy degradation. Our non-uniform quantization approach provides a novel alternative to the existing uniform quantization techniques for neural networks. We further propose a novel complexity metric of number of bit operations performed (BOPs), and we show that this metric has a linear relation with logic utilization and power. We suggest evaluating the trade-off of accuracy vs. complexity (BOPs). The proposed method, when evaluated on ResNet18/34/50 and MobileNet on ImageNet, outperforms the prior state of the art both in the low-complexity regime and the high accuracy regime. We demonstrate the practical applicability of this approach, by implementing our non-uniformly quantized CNN on FPGA.